kind suggestion for an high current buffer.

[beppe61]

Dear Sir,


I would like to built an high current buffer capble of sourcing let's say at least 5A on a 4 ohm load.
I need it to build a complete amp of about 100 W on 4 ohm.
All the needed voltage gain will be realized in a separate stage.
The requirements for the buffer are:
1) just a pair of final BJTs and in general a minimum number of components
2) a reasonably high input impedance.
It should drive easily loads of 4 ohm..
This idea follows that of an amp that I listened some years ago where all the voltage gain was realized with tubes and the current gain with Sanken BJTs.
I have already a nice solid state voltage gain stage schematic to use but as you can understand I lack a suitable current gain stage schematic.
So I am looking for suitable schematics.
The voltage gain stage and the current gain stage will be capacitor coupled.

Thank you very much indeed.

Kind regards,

beppe61

 

[PMA]

I have one on my web page (push-pull follower).

http://www.pha.inecnet.cz/macura/next.htm

 

[joensd]

Here are some very interesting designs to take into consideration, although they wouldn´t exactly meet your design goals, like one output pair etc.
http://www.audiodesignguide.com/my/
I tried the "Single Ended Pure Class A Power Follower Amplifier" at one point and it is very pure indeed.
See also hot follower or cold follower if you don´t want class a.

Just curious which gain stage did you decide to use?
greets

 

[lumanauw]

I'm also looking for follower, but from classD type. It has no crossover distortion, runs cool, and I can put tube or any high THD voltage swing infront of it
Anyone have seen this kind of follower?

 

[PMA]

class D - they are fine. The only problem is to listen to them. A colleague of mine develops them and faces pretty listening surprises.

 

[CBS240]

Dear Sir,


I would like to built an high current buffer capble of sourcing let's say at least 5A on a 4 ohm load.
I need it to build a complete amp of about 100 W on 4 ohm.
All the needed voltage gain will be realized in a separate stage.
The requirements for the buffer are:
1) just a pair of final BJTs and in general a minimum number of components
2) a reasonably high input impedance.
It should drive easily loads of 4 ohm..
This idea follows that of an amp that I listened some years ago where all the voltage gain was realized with tubes and the current gain with Sanken BJTs.
I have already a nice solid state voltage gain stage schematic to use but as you can understand I lack a suitable current gain stage schematic.
So I am looking for suitable schematics.
The voltage gain stage and the current gain stage will be capacitor coupled.

Thank you very much indeed.

Kind regards,

beppe61

Hi beppe61,


You should be able to get this kind of current out of 1 pair, no problem. However, the input impeadence will depend on the Hfe of the output transistors. I think you should include this stage inside the feedback loop, if possible, DC. Well if you use any GNF. If you use a darlington EF circuit to increase Hfe, remember to use emitter degeneration resistors and that you might need thermal compesation...like a Vbe multiplier or something. You might also use a couple of current sources for biasing. Is this an addition on an existing amp or for a new circuit?

 

[lumanauw]

class D - they are fine. The only problem is to listen to them. A colleague of mine develops them and faces pretty listening surprises.

PMA, if you say it is "problem", I got a feeling that the "surprises" is about not good sounding. Common comment is "sterile" or "has no life sound". Inherently, all classD (with global feedback) will be eliminating all distortions (or transfering them to all high order ones, seen as "noise floor", due to HF switching artifacts is re-entering the differential).

It is hard to tailor the harmonic pattern weigh decay (bigger smaller order ones, decaying towards high order ones) in classD, they tend to jump directly to high order ones. This makes the sound is "typical" of classD.

This pattern is important, independent of the THD number. No matter 0.5 or 0.005, it is the harmonic pattern that matters

Usually the so called classD amp is a complete amp, where the comparator/differential, the voltage swing and the current section is coming from the same engine, global feedback, the whole amp is complete classD.

Do you think it will be different sound if we only use classD limited to the current section, while the comparator/differential and the voltage swing sections are made from very linear analog type / classA or even tube section?

I also think to make the differential+voltage swing section to be independent section of the whole amp, not having any feedback from the output node. But this needs a current section that has somekind of local loop to be independent, yet very linear input-output locally. (has the characteristic your NP-PMA, very linear input-output, able to stand alone without global feedback)

This way the "classD" characteristic will be donating very small portion of the whole sound, because the voltage swing is made by someone else, the classD is only follower, but we can get 2 of its major attrataction, very cool and not having crossover distortion.

 

[beppe61]

Re: Re: kind suggestion for an high current buffer.

Hi beppe61,
You should be able to get this kind of current out of 1 pair, no problem. However, the input impeadence will depend on the Hfe of the output transistors. I think you should include this stage inside the feedback loop, if possible, DC. Well if you use any GNF. If you use a darlington EF circuit to increase Hfe, remember to use emitter degeneration resistors and that you might need thermal compesation...like a Vbe multiplier or something. You might also use a couple of current sources for biasing. Is this an addition on an existing amp or for a new circuit?

Dear Sir,
thank you very much for your kind reply.
I would like to explain my idea.
Some years ago I had the opportunity to listen to an "hybrid" amp
of about 50 W where the gain stage was tubed and the final current stage was solid state.
I have some schematic for solid state voltage gain stage (up to 6X).
So I need a nice and simple current gain stage to couple with.
As I am minimalist by wallet, a like very much simple topologies built around cheap components.
Not always cheap means bad.
I know of well regarded amps that sport the cheap 2N3055 and they do sound well.
In particular the 2N3055 would be a component of choice.

Thank you very much again.

Kind regards,

beppe61

 

[beppe61]

I have one on my web page (push-pull follower).
http://www.pha.inecnet.cz/macura/next.htm

Thank you very much Sir.
I am looking for something even more basic, something like 4 bjts only.
But I am not sure it is feasible.

Kind regards,

beppe61

 

[beppe61]

Here are some very interesting designs to take into consideration, although they wouldn´t exactly meet your design goals, like one output pair etc.
http://www.audiodesignguide.com/my/
I tried the "Single Ended Pure Class A Power Follower Amplifier" at one point and it is very pure indeed.
See also hot follower or cold follower if you don´t want class a.
Just curious which gain stage did you decide to use?
greets

Dear Sir,
thank you very much for the kind and valuable reply.
I like in particular the schematic attached.
That is qhat I am looking for (maybe with just one pair output devices).
I would like to test two or three voltage gain stage.
2 are solid state (Bride of Zen, schematic taken by an old preamp Radford ZD22) and 1 tube (my friend OTL headphone amplifier).
The buffer should do the dirty work of providing the high current to drive my reluctant Dynaudio speakers.

Kind regards,

beppe61

 

[CBS240]

Re: Re: Re: kind suggestion for an high current buffer.

I have some schematic for solid state voltage gain stage (up to 6X).
So I need a nice and simple current gain stage to couple with.
As I am minimalist by wallet, a like very much simple topologies built around cheap components.
Not always cheap means bad.
I know of well regarded amps that sport the cheap 2N3055 and they do sound well.
In particular the 2N3055 would be a component of choice.

Thank you very much again.

Kind regards,

beppe61

Any idea of the output impeadence of the gain stage you plan to use? The beta (Hfe) of a 2N3055 drops after a few amps, and becomes non-linear. This might create even more distortion if not included in the feedback loop. There are more linear devices available even in matched pairs specially designed for audio. Not too expensive and probably worth the result. I'd shoot for more like 60 quality W from 1 pair of 2N3055/MJ2955 on 4Ohms.

Mj21193/4 are pretty rugged devices from On-Semi. This would be a better 4Ohm driver. Big SOA and harder to destroy. Compare this device with a 2N3055 and you will see the difference.

 

[beppe61]

Re: Re: Re: Re: kind suggestion for an high current buffer.

Thank you Sir for your again kind reply.

Any idea of the output impeadence of the gain stage you plan to use?

Worst case should be 1 kohm, capacitor coupled with the high currente buffer stage.
I really don't know if having also a 2X voltage gain from the buffer could complicated it too much.
Thank you so much.

Kind regards,

beppe61